Must read from the New York Times.
Published: January 24, 2011
Books: A Pound of Prevention Is Worth a Closer Look (January 25, 2011)
Our Enthusiasm for Diagnosis
My first car was a ’65 Ford Fairlane wagon. It was a fairly simple—albeit large—vehicle. I could even do some of the work on it myself. There was a lot of room under the hood and few electronics. The only engine sensors were a temperature gauge and an oil-pressure gauge.
Things are very different with my ’99 Volvo. There’s no extra room under the hood—and there are lots of electronics. And then there are all those little warning lights sensing so many different aspects of my car’s function that they have to be connected to an internal computer to determine what’s wrong.
Cars have undoubtedly improved over my lifetime. They are safer, more comfortable, and more reliable. The engineering is better. But I’m not sure these improvements have much to do with all those little warning lights.
Check-engine lights—red flags that indicate something may be wrong with the vehicle—are getting pretty sophisticated. These sensors can identify abnormalities long before the vehicle’s performance is affected. They are making early diagnoses.
Maybe your check-engine lights have been very useful. Maybe one of them led you to do something important (like add oil) that prevented a much bigger problem later on.
Or maybe you have had the opposite experience.
Check-engine lights can also create problems. Sometimes they are false alarms (whenever I drive over a big bump, one goes off warning me that something’s wrong with my coolant system). Often the lights are in response to a real abnormality, but not one that is especially important (my favorite is the sensor that lights up when it recognizes that another sensor is not sensing). Recently, my mechanic confided to me that many of the lights should probably be ignored.
Maybe you have decided to ignore these sensors yourself. Or maybe you’ve taken your car in for service and the mechanic has simply reset them and told you to wait and see if they come on again.
Or maybe you have had the unfortunate experience of paying for an unnecessary repair, or a series of unnecessary repairs. And maybe you have been one of the unfortunate few whose cars were worse off for the efforts.
If so, you already have some feel for the problem of overdiagnosis.
I don’t know what the net effect of all these lights has been. Maybe they have done more good than harm. Maybe they have done more harm than good. But I do know there’s little doubt about their effect on the automotive repair business: they have led to a lot of extra visits to the shop.
And I know that if we doctors look at you hard enough, chances are we’ll find out that one of your check-engine lights is on.
A routine checkup
I probably have a few check-engine lights on myself. I’m a male in my mid-fifties. I have not seen a doctor for a routine checkup since I was a child. I’m not bragging, and I’m not suggesting that this is a path others should follow. But because I have been blessed with excellent health, it’s kind of hard to argue that I have missed out on some indispensable service.
Of course, as a doctor, I see doctors every day. Many of them are my friends (or at least they were before they learned about this book). And I can imagine some of the diagnoses I could accumulate if I were a patient in any of their clinics (or in my own, for that matter):
• From time to time my blood pressure runs a little high. This is particularly true when I measure it at work (where blood pressure machines are readily available).
Diagnosis: borderline hypertension
• I’m six foot four and weigh 205 pounds; my body mass index (BMI) is 25. (A “normal” BMI ranges from 20 to 24.9.)
• Occasionally, I’ll get an intense burning sensation in my midchest after eating or drinking. (Apple juice and apple cider are particularly problematic for me.)
Diagnosis: gastroesophageal reflux disease
• I often wake up once a night and need to go to the bathroom.
Diagnosis: benign prostatic hyperplasia
• I wake up in the morning with stiff joints and it takes me a while to loosen up.
Diagnosis: degenerative joint disease
• My hands get cold. Really cold. It’s a big problem when I’m skiing or snowshoeing, but it also happens in the office (just ask my patients). Coffee makes it worse; alcohol makes it better.
Diagnosis: Raynaud’s disease
• I have to make lists to remember things I need to do. I often forget people’s names—particularly my students’. I have to write down all my PINs and passwords (if anyone needs them, they are on my computer).
Diagnosis: early cognitive impairment
• In my house, mugs belong on one shelf, glasses on another. My wife doesn’t understand this, so I have to repair the situation whenever she unloads the dishwasher. (My daughter doesn’t empty the dishwasher, but that’s a different topic.) I have separate containers for my work socks, running socks, and winter socks, all of which must be paired before they are put away. (There are considerably more examples like this that you don’t want to know about.)
Diagnosis: obsessive-compulsive disorder
Okay. I admit I’ve taken a little literary license here. I don’t think anyone would have given me the psychiatric diagnoses (at least, not anyone outside of my immediate family). But the first few diagnoses are possible to make based solely on a careful interview and some simple measurements (for example, height, weight, and blood pressure).
More are possible if a doctor were to order any one of a number of diagnostic tests for me. Even routine blood work—a complete blood count, an electrolyte panel, and liver function tests—involves more than twenty separate measurements. The chances are good that I would have at least one abnormal value.
And then there’s imaging. Lots of people have “abnormal” findings on X-ray studies. If I had a chest X-ray, I wouldn’t be surprised if a lung nodule was seen. If I had an abdominal CT scan, I wouldn’t be surprised if a cyst on my kidney was found.
Further inspection could reveal more. A colonoscopy might show that I have polyps—as about a third of people my age do. A prostate biopsy might demonstrate a small cancer—which many men have, even if their PSA (prostate-specific antigen) screening tests are normal. And it’s a safe bet that my genome contains all sorts of genetic variants.
To be fair, most doctors wouldn’t order any imaging studies; some might have skipped the routine blood work. Nonetheless, several of these diagnoses could have been made.
Would I be better off if I were given these diagnoses? I don’t think so. Would I be put on prescription medications? Probably. Would I consider this good medical care or bad? I’d say bad. But enough about me. This book is about the millions of Americans who have access to what some would call the best medical care in the world. Of course, there are millions of other Americans whose access is severely limited—the uninsured. This is a real problem, but not the topic of this book. The problems described here are actually less likely to happen to the latter group, simply because they receive less medical care. This book is about the relentless expansion of medicine and our increasing tendency to make diagnoses.
Americans have been trained to be concerned about our health. All sorts of hidden dangers lurk inside of us. The conventional wisdom is that it’s always better to know about these dangers so that something can be done. And the earlier we know, the better. That’s why we are so enthusiastic about amazing medical technologies that can detect abnormalities even when we think we are well. That’s also why we welcome the identification of risk factors, disease awareness campaigns, cancer screening, and genetic testing. Americans love diagnosis, especially early diagnosis.
Not surprisingly, we get more diagnoses today than we did in the past. In fact, we are in the midst of an epidemic of diagnosis. Again, the conventional wisdom tells us that this is good: finding problems early saves lives because we have the opportunity to fix small problems before they become big ones. What’s more, we believe there are no downsides to looking for things to be wrong.
But the truth is that early diagnosis is a double-edged sword. While it has the potential to help some, it always has a hidden danger: overdiagnosis—the detection of abnormalities that are not destined to ever bother us.
Living longer, yet sicker?
Consider the generation of which I am a part—the baby boomers. This is the generation born in the period of increased birthrates that followed World War II. They went on to become leaders in the major social movements of the 1960s—civil rights, feminism, and the Vietnam War protests. They also spawned the counterculture of that era: sex, drugs, and rock and roll. As they aged, they became the dominant culture: they gained political power and amassed large sums of money. Now television ads promise them that they will engage in a new kind of retirement, one in which their dreams won’t retire. Just think of the Ameriprise ad featuring the late Dennis Hopper saying, “’Cause I just don’t see you playing shuffleboard—know what I mean?” (while the powerhouse organ riff from the classic rock hit “Gimme Some Lovin’” blares in the background). Brings back fond memories of high school. I love it.
But then I saw a piece in the Washington Post suggesting that the boomers might indeed need to prepare for a different view of retirement—because they are falling apart.1 Large national surveys reported that while 57 percent of those born before World War II reported excellent health as they approached retirement, only 50 percent of boomers described themselves in this way. About 56 percent of those born before World War II reported having a chronic condition at retirement; about 63 percent of boomers reported having a chronic condition at the same age. Could boomers be in worse shape than their parents were?
A few weeks later I attended a medical meeting at which one of the participants reported on the Department of Health and Human Services’ midcourse review of the program called Healthy People 2010. This is the federal government’s effort to increase both the length and the quality of life. Length of life was measured using life expectancy—the average number of years Americans live. Quality of life was measured using healthy life expectancy—the average number of years Americans live free of disease (such as heart disease, stroke, cancer, diabetes, hypertension, and arthritis). The speaker showed a table with data from 1999 to 2002, during which life expectancy had increased by about six months, from 76.8 to 77.2 years. But surprisingly, the healthy life expectancy had fallen by a little more than a year, from 48.7 to 47.5.
It looked like the program was getting it only half right: the quantity of life was increasing (people were living longer), but the length of healthy life was decreasing (people were having fewer disease-free years). Could we be living longer, yet be sicker? That is hard to believe. But there is an alternative explanation: we live longer, we are healthier, but we are increasingly more likely to be told we are sick.
Some may view diagnosing more people (and treating more people) as the price that has to be paid for most of us to achieve an extension of life. This assumes that early diagnosis and treatment is the only explanation of a longer life span. But because other things are more important (such as not smoking, nutrition, exercise, and medical care for the acutely ill), it’s likely that most of this life extension would occur regardless of whether or not there was more diagnosis. And since for many, length of life is not the only goal, questions about whether the health-care system introduces disease and disability into the population become more relevant.
What this book is about
My mother thinks she knows what this book is about. She is almost ninety and has advanced dementia. A few months ago she picked up my first book and read the title out loud: “Should I Be Tested for Cancer?” And then she answered with a resounding “No!” (Note: that title is a vast oversimplification of the book’s content.)
She asked me what my next book would be about. I attempted to explain it to her. She suggested that it be titled Should I Be Tested for Anything? Not that great a title, but it gives you the idea. This book examines the possibility that American medicine now labels too many of us as “sick.”
As I’ve noted, the conventional wisdom is that more diagnosis—particularly, more early diagnosis—means better medical care. The logic goes something like this: more diagnosis means more treatment, and more treatment means better health. This may be true for some. But there is another side to the story. More diagnosis may make healthy people feel more vulnerable—and, ironically, less healthy. In other words, excessive diagnosis can literally make you feel sick. And more diagnosis leads to excessive treatment—treatment for problems that either aren’t that bothersome or aren’t bothersome at all. Excessive treatment, of course, can really hurt you. Excessive diagnosis may lead to treatment that is worse than the disease.
More specifically, this book is about overdiagnosis. While the term sounds like it means simply “excessive diagnosis,” it actually also has a more precise meaning. Overdiagnosis occurs when individuals are diagnosed with conditions that will never cause symptoms or death.
So while I diagnosed myself with a number of conditions a few pages ago, some were not overdiagnoses, since I had symptoms: heartburn, cold hands, and so forth (although they may well constitute excessive diagnoses, given that my symptoms were trivial). But the diagnoses related to slight elevations in blood pressure and weight were not associated with symptoms. They could reflect overdiagnosis. So too could all the diagnoses I might have gotten following subsequent testing. In other words, overdiagnosis can occur only when a doctor makes a diagnosis in a person who has no symptoms referable to the condition. While this can happen when a doctor stumbles onto unexpected diagnoses in the course of an evaluation of unrelated conditions, generally it happens because doctors seek early diagnoses—either as part of an organized screening effort or during routine exams. Thus, overdiagnosis is a consequence of the enthusiasm for early diagnosis.
The trouble is that we doctors don’t know if an individual has been overdiagnosed unless that person forgoes treatment, lives the rest of his or her life symptom free, and dies from some other cause. But we do know that if we make more and more diagnoses in a healthy population, we are more likely to overdiagnose.
Overdiagnosis is a relatively new problem in medicine. In the past, people didn’t go to the doctor when they were well—they tended to wait until they developed symptoms. Furthermore, doctors didn’t encourage the healthy to seek care. The net result was that doctors made fewer diagnoses than they do now.
But the paradigm has changed. Early diagnosis is the goal. People seek care when they are well. Doctors try to detect disease earlier. More people have findings of early disease than of late disease. So we make more diagnoses—including diagnoses in those who have no symptoms. Some of these people are destined to develop symptoms. Others are not—they are overdiagnosed.
So the problem of overdiagnosis stems directly from the expansion of the pool of individuals in whom we make diagnoses: from individuals with disease (those with symptoms) to individuals with abnormalities (those without symptoms). The problem is further aggravated as the definition of what constitutes an abnormality gets increasingly broad.
The objective of this book is to lay out the data on how overdiagnosis occurs, explain why it can be harmful, and explore its root causes. My hope is to help you critically consider the desirability of being turned into a patient prematurely.
Let me be clear about why you should care about overdiagnosis. Since doctors don’t know who is overdiagnosed and who is not, overdiagnosed patients tend to get treated. But an overdiagnosed patient cannot benefit from treatment. There’s nothing to be fixed—he will neither develop symptoms nor die from his condition—so the treatment is unneeded. An overdiagnosed patient can only be harmed. And the simple truth is that almost all treatments have the potential to do some harm.
What this book is not about
This book is not about what you should do when you are sick. It is not for the few who are severely ill (those for whom medical care offers a lot), but for the many who are (or used to be) basically well—or those who have one illness and are at risk of being told they have others. Nor is this book an apology for sloppy diagnosis in the sick. Diagnosis is always important when people are suffering, and it’s important that it be done well. None of my comments should be construed as suggesting you are better off not being diagnosed when you are sick. Finally, this book is not a condemnation of all of American medicine, nor a call for alternative medicine. I’m conventionally trained in Western medicine, and I believe doctors do a lot of good. If you are sick, you should see one.
A final note about people and language
Before moving on, I feel obliged to make a few comments about names and words. There are stories in this book: stories about my patients, my friends, and people I’ve met along the way. The stories are accurate; the names are not. While I have not altered information relevant to the clinical narrative (such as the individual’s gender, age, symptoms, and experiences), I have altered information that could potentially identify individuals (such as whether a person is from New York or New Jersey—as my daughter might say, “Like it matters”).
Then there is the word disease. Although the word has a wide range of interpretations, its origins are quite specific. Dis- means “without,” and ease requires no explanation. A synonym for disease might be discomfort. Although there are other perfectly legitimate definitions, in this book disease will refer to a condition that a person experiences—a sickness, an illness, a disorder that produces symptoms.
The word abnormality will serve a distinct purpose. I will use it to describe findings that are considered abnormal in the medical profession yet are not experienced by the individual. Some of the most familiar abnormalities—high blood pressure, high cholesterol—will sometimes be referred to as conditions to distinguish them from diseases.
Although occasionally I use the broad term health-care provider, for simplicity, I tend to use the term doctor. This is not meant to exclude other caregivers. On the contrary, it is important to acknowledge that physician assistants and nurse-practitioners are assuming larger and more important roles in medicine—particularly in the delivery of primary care (where a lot of diagnoses are made).
Finally, some quick notes about pronouns. The most familiar are he and she. Of course, a patient can be either male or female, as can a doctor (there are now more women than men enrolled at Dartmouth’s medical school). I don’t know of a satisfactory way to handle the absence of a gender-neutral singular pronoun. He or she gets pretty awkward after a while; using they would upset my mother too much. So when the situation allows (some diseases are gender specific), I alternate between the two.
Then there’s we. We will generally refer to “we doctors” or “we health-care providers.” (I’d guess generally means “roughly 90 percent of the time”—although I’m not going to bother to calculate it.) I use we in an attempt to represent the professional perspective of doctors: what we are taught in medical school, how we are trained as residents, what we learn in practice. In short, I’ll try to give you a sense of how we think. Not that we all think alike, but we do all share a common experience, about which you should have some insight.
Occasionally we will refer to “we the public.” Just like you, I am a member of society and a potential patient. And all of us will face some decision about how we want to interact with medical care. Sometimes I modify the we with something like “we the public” when I am attempting to communicate this perspective.
I will represent me—the author. But it should be another we, as this book is really a collaboration of three authors: Dr. Lisa Schwartz, Dr. Steven Woloshin, and myself. But to avoid the confusion with the other wes requires this sleight of hand. To be clear, our voice encompasses two viewpoints. All three of us are academic physicians: we see patients, we teach students, and we do research. But we are also people, and therefore potential patients. As people, we are concerned about the relentless expansion of the medical profession and the subsequent drive to turn people into patients. It is the melding of these two viewpoints—medical and personal—that provides the motivation for this book.
Excerpted from Overdiagnosed: Making People Sick in the Pursuit of Health by Dr. H. Gilbert Welch, Dr. Lisa M. Schwartz, and Dr. Steven Woloshin. Copyright 2011. Excerpted with permission by Beacon Press