Uh-oh, it looks like your Internet Explorer is out of date.
For a better shopping experience, please upgrade now.
Getting tested to detect cancer early is one of the best ways to stay healthyor is it?
In this lively, carefully researched book, a nationally recognized expert on early cancer detection challenges one of medicine's most widely accepted beliefs: that the best defense against cancer is to always try to catch it early. Read this book and you will think twice about common cancer screening tests such as total body scans, mammograms, and prostate-specific antigen (PSA) tests.
Combining patient stories and solid data on common cancers, Dr. H. Gilbert Welch makes the case that testing healthy people for cancer is really a double-edged sword: while these tests may help, they often have surprisingly little effect and are sometimes even harmful. Bringing together a body of little-known medical research in an engaging and accessible style, he discusses in detail the pitfalls of screening tests, showing how they can miss some cancers, how they can lead to invasive, unnecessary treatments, and how they can distract doctors from other important issues. Welch's conclusions are powerful, counterintuitive, and disturbing: the early detection of cancer does not always save lives, it can be hard to know who really has early cancer, and there are some cancers better left undiscovered.
Should I Be Tested for Cancer? is the only book to clearly and simply lay out the pros and cons of cancer testing for the general public. It is indispensable reading for the millions of Americans who repeatedly face screening tests and who want to make better-informed decisions about their own health care.
About the Author
H. Gilbert Welch, M.D., M.P.H., is Professor in the Departments of Medicine and Community and Family Medicine at Dartmouth Medical School and Codirector of the VA Outcomes Group in the Department of Veterans Affairs, White River Junction, Vermont.
Read an Excerpt
SHOULD I BE TESTED FOR CANCER?MAYBE NOT AND HERE'S WHY
By H. Gilbert Welch
The University of California PressCopyright © 2004 Regents of the University of California
All right reserved.
Chapter OneYou may have a "cancer scare" and face an endless cycle of testing
In our society, information gathering is viewed almost uniformly as a good thing. (It is the "information age," after all.) Nowhere is this more true than in medicine. For doctors, more information is always better. In the past, most of our information came from the patient. Now it increasingly comes from machines.
Doctors like tests because we see them as objective and more reliable than our own subjective judgments. We also see tests as something tangible we can offer the patient at the end of a clinic visit. Patients like tests for the same reasons. Ordering a test validates their concerns and promises concrete information-a definitive diagnosis. Sometimes patients even perceive their care as substandard if they are not given some sort of test. While doctors and patients recognize that treatments may have side effects or lead to complications, both tend to view testing as something that can only help. The prevailing attitude seems to be It can't hurt just to gather a little information.
Of course, that is not always true. In this chapter I describe the most familiar problem with cancer testing: the test can be wrong. In short, people with abnormal screening tests often don't have cancer. But before they find out for sure, they may have to go through multiple tests-tests that may be unpleasant and that may lead to complications. Throughout the testing period, they will worry about whether they have cancer. And some may never get a definitive answer. That can hurt a lot.
TRAPPED IN AN ENDLESS CYCLE OF TESTING
Every other week I see patients in the Veterans Administration walk-in clinic, a clinic for patients who either don't have an appointment or don't have a doctor. I recently saw a gentleman who wanted to have his cholesterol tested. He also wanted to talk about PSA screening for prostate cancer. Like many of our veterans, he came into the examination room accompanied by his wife. Both were in their early 70s.
His cholesterol was fine. I asked him what he'd like to know about prostate cancer screening, and he said he wanted to know why there was any debate as to its usefulness. Just by knowing there was a debate, he was further along than many. I said we really didn't know whether it saved lives or not. He said, "What's the harm in trying?" I told him that many older men have elevated PSAs (because of enlarged prostate glands) and yet don't have prostate cancer. But the only way they can find out that they don't have cancer is to have a prostate biopsy (a procedure no one enjoys-more on that later). I also mentioned that some people have indeterminate biopsy results: they won't be told they have cancer, but they also won't be told they do not.
His wife, who had been listening quietly, now spoke up: "It's like ASCUS, isn't it?"
I was stunned. I was talking about the ambiguities of PSA screening for prostate cancer; she thought immediately of Pap smear screening for cervical cancer. It was a remarkable connection, one many doctors might miss. And she was right. ASCUS stands for "a typical squamous cells of unknown significance," which are frequently detected in a Pap smear. It's not cancer, but it's not normal. Instead we call it an "indeterminate" result. I told her the analogy was right on target.
She and her husband then shared what she had been through over the past five years. She had been told that her Pap smear was abnormal. It was then repeated every three to six months. She had had a colposcopy: an optical instrument was placed in the vagina to better visualize the cervix. She had been biopsied. She had been told she did not have cancer. But the Pap smears were still abnormal. She had had cryocauterization: a cold probe was used to freeze and kill cells on the cervix. She had had laser therapy: a high energy light beam was used to burn and kill cervical cells. Most recently she had had a cervical conization: a procedure in which the core of the cervix is cut out. Her Pap smear was still abnormal.
Some doctors were suggesting that she have a hysterectomy-removal of her entire uterus and cervix-even though they could not prove she had cancer. Others were saying she should just keep checking. Neither approach appealed to her. She was fed up. What she wanted was the answer to an apparently simple question: "Is there a problem or not?"
TESTS ARE IMPERFECT
In an ideal world, we'd have ideal tests. They'd be cheap, simple, safe, and quick. And they would never be wrong. Among people who had cancer, the test for cancer would always be positive, while among those who didn't have cancer, the test would always be negative. Positive tests could immediately be followed by early treatment; negative tests would result in immediate reassurance. This utopian ideal would look something like this:
Cancer, however, is a diagnosis made by examining human tissue under the microscope. And the only way to look at tissue under the microscope is to do a biopsy: cut a small piece of tissue and remove it from the body. A biopsy is a small operation, and like any operation, it can be disruptive and painful and can lead to complications. So it's not the kind of test you want to perform on everyone.
The job of the cancer screening test is to determine which patients should be biopsied. In other words, a screening test is a preliminary test. It is not a test to determine who has cancer; instead, it is a test to determine who should be tested further. So the more pragmatic ideal looks like this:
As you may infer from the illustration, a positive screening test can be wrong. That is, a person can be told that his screening test suggests the possibility of cancer, while the biopsy demonstrates that no cancer exists. A positive screening test that is proved wrong is called a false positive. Many call it a "cancer scare."
Can a negative screening test be wrong? The answer is almost certainly yes, although it is very hard to prove. That is because we do not biopsy people with negative screening tests. The only way we ever come to suspect that a negative screening test might have been wrong is when a new cancer becomes clinically obvious soon after a person has a negative test. Recall from the previous chapter the story about my patient's daughter who had metastatic breast cancer diagnosed three months after a normal mammogram. In situations like this, it is reasonable to wonder whether the normal result-or negative screening-was wrong. But as I suggested then, the problem could just as easily be a fast-growing cancer as a falsely negative test. It's impossible to know.
TESTING IN THE REAL WORLD
In the real world, cancer testing is more complex. Test results aren't just positive or negative; often they are somewhere in between. These in-between results may lead to in-between tests: something more thorough than the screening test but a few steps short of a biopsy. Sometimes in-between results lead to a recommendation simply to repeat the screening test after a few months have passed. The reality, therefore, looks more like this:
Most people will be on the first negative arrow following a screening test: they will avoid this complex cycle altogether. At the same time, a lot more people will be on the positive or in-between arrows than will ever have cancer. False positive and indeterminate results are inherent features of screening-because unfortunately, we don't have perfect tests.
The woman I told you about at the beginning of this chapter has been on every arrow in this figure, except those leading to a definitive diagnosis. You might think of it as a testing cascade, with one test begetting another. Alternatively, you might think of it as being trapped in an endless cycle, where the patient keeps coming back for more testing. Either way it's no fun. What happened to her may not happen often, but it does happen.
Although the testing process differs depending on which cancer is being sought, some features are common to all cancer testing. The screening test itself is generally the simplest, least disruptive, and safest of all the tests used to detect a cancer. When the screening test is abnormal, confirmatory testing is initiated. The confirmatory testing can range from repeating the screening test earlier than normal (in six months instead of one year, for example) to performing an alternative test that is generally more accurate (and generally more involved) to performing a biopsy (the most definitive test). If the screening test is not too suspicious, an early repeat test or an alternative test is often recommended. But if it is suspicious, a biopsy is generally recommended. The confirmatory testing options for common screening tests are shown in Table 2.
HOW COMMON ARE FALSE POSITIVE TESTS?
Now that you know something about the testing process, I want to give you some sense of how often problems occur. I'll start with the false positive test, or "cancer scare": a positive screening test that, after further confirmatory testing, is determined to be in error. A false positive triggers a process during which people can get hurt, in terms of both their physical and mental health.
The measure of how often a screening test is falsely positive is called a false positive rate: if 100 people are screened and 5 have false positive results, the false positive rate is 5 percent. The false positive rate depends on a number of factors: the test itself, the population being tested, the quality of the testing, and how the test is interpreted. All the common screening tests are plagued by the problem of false positive results.
Investigators in Europe have reported false positive rates of about 5 percent for women undergoing mammography for the first time. Women receiving subsequent mammograms were about half as likely to experience a false positive. This is the general pattern with mammography. Here's why: with the first mammogram, the radiologist has no other picture for comparison. All breasts are different, and it is not always clear what is normally different versus what is abnormally different. Without a previous mammogram for reference, therefore, radiologists understandably play it conservative and read more films as abnormal. But with subsequent mammograms there is a point of comparison-an earlier picture of the same breast. Now all the radiologist has to do is look for changes. Something that appeared worrisome in one picture yet is the same in the next becomes a lot less worrisome. So there tend to be fewer false positives on subsequent mammograms.
Investigators at the University of California, San Francisco-who devote particular effort to doing mammography well-report slightly higher false positive rates for first-time exams: about 6 to 7 percent. Following the normal pattern, their subsequent exams then have lower false positive rates. But this is an exceptionally good practice; for the rest of the United States, it appears, the experience is quite different. Some investigators (including myself) have measured what goes on where most patients receive their health care: a typical community practice. Despite the standard mixture of first-time and subsequent testing, the false positive rate for most American women seems to be much closer to 10 percent. The reason? No one knows for sure, but I suspect that most American radiologists err on the side of calling things abnormal (in part, for fear of being sued), which in turn leads to higher false positive rates.
Prostate specific antigen (PSA)
Now let's consider the blood test for prostate cancer, the prostate specific antigen, or PSA. In September 1993, during "Prostate Cancer Awareness Week," a major effort was made to recruit volunteers for testing. Of the more than 30,000 men over age 50 tested in 148 centers, about 3,000 had an abnormally elevated level of PSA; that is, about 10 percent tested positively. Somewhere between a quarter and a third of these men were found to have prostate cancer, making the false positive rate about 7 percent. Others have reported false positive rates in the range of 7 to 11 percent. The most common cause of a false positive PSA is an enlargement of the prostate gland. Because the prostate tends to enlarge with age, PSAs tend to rise with age. Therefore, false positives are less common in younger men and more common in older men.
The false positive rate for subsequent PSA testing depends on whether the initial result was normal or not. If the men with abnormal PSAs are excluded, the false positive rate for a repeat test drops to around 2 percent. In other words, if your test was normal in the past, you are much less likely to have a falsely abnormal test in the future (which in turn means that an abnormal result in subsequent testing is more likely to represent a cancer). However, if your PSA was abnormally elevated once, it is very likely to remain abnormal in the future-even if you never have prostate cancer.
Fecal occult blood testing
Fecal occult blood testing (in which stool is collected and tested for blood as an early sign of colon cancer) has perhaps the highest false positive rate of any screening test, between 8 and 16 percent. The reason is that there are many other possible sources of blood: irritation of the stomach, ulcers in the intestine, and hemorrhoids in the rectum, to name a few. The false positive rate is high enough that many doctors try to prepare patients for a false positive at the time of testing. I tend to say something like, "If your test is positive, it doesn't mean you have cancer-it just means we need to look further and see where the blood is coming from." The false positive rate is about the same on subsequent exams.
It is more difficult to talk about the false positive rate for Pap smears because there are two types of abnormal results. First, roughly 10 percent of Pap smears need to be repeated because of inflammatory changes, generally related to infection. Here, age is again an important variable: adolescents have a false positive rate of more than 15 percent, whereas for women over 60 it is less than 5 percent.
Second, around 5 percent of women undergoing a Pap smear will be told they have a cellular abnormality that is potentially worrisome for cancer. This number also varies by age. Among adolescents the proportion of smears diagnosed with ASCUS (atypical squamous cells of unknown significance) or SIL (squamous intraepithelial lesion) has been reported to be as high as 14 percent. Since almost no women with ASCUS and SIL have or will develop invasive cervical cancer, virtually all of these test results can be considered false positives. However, many of these abnormalities end up being treated, so the problem can also be characterized as unnecessary diagnosis.
Excerpted from SHOULD I BE TESTED FOR CANCER? by H. Gilbert Welch Copyright © 2004 by Regents of the University of California. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
|Introduction: The conventional wisdom about cancer testing and what this book is about||1|
|Part I||Problems You Should Know about|
|1.||It is unlikely that you will benefit||17|
|2.||You may have a "cancer scare" and face an endless cycle of testing||33|
|3.||You may receive unnecessary treatment||51|
|4.||You may find a cancer you would rather not know about||66|
|5.||Your pathologist may say it's cancer, while others say it's not||90|
|6.||Your doctor may get distracted from other issues that are more important to you||106|
|Part II||Becoming a Better-Educated Consumer|
|7.||Understand the culture of medicine (and why we are pushed to test)||115|
|8.||Understand the statistics of cancer (and why five-year survival is the world's most misleading number)||129|
|9.||Understand the limits to research--even genetic research (and why it is hard to be sure there really are benefits to screening)||152|
|10.||Develop a strategy that works for you||177|
|Appendix||Summary of cancers discussed in this book||191|
What People are Saying About This
"One of those gems to come out of the academic press. . . . If you're worried about cancer, this lucidly argued book will be a godsend."The Week