To Your Health
By James N. Dillard, M.D.
(11/26/2009) We can all be thankful that we live in a country where experts and authorities can change their minds when they need to. Flexibility is good, but change that may affect our very lives can be anxiety-provoking. Doctors and public health leaders have spent years reinforcing the need for mammograms and Pap smears at certain ages, but now the recommendations are changing.
Last week, the Preventive Services Task Force of the Department of Health and Human Services recommended that screening mammograms begin at age 50 instead of age 40, and that they be performed only every other year. The federally appointed expert panel questioned the usefulness of mammography after age 75, and it recommended against breast self-examinations.
In the same vein, the American College of Obstetricians and Gynecologists is about to loosen its recommendations for first Pap smear age and subsequent frequency. The print media and television have made much ado about these changes, and people seem upset. The turmoil may be mostly because it is challenging to understand the fundamental concepts of medical screening tests.
Health screening tests are only about looking for hidden or impending disease (doctors call it clinically occult disease). We are all at risk for illness, and we’d like to be able to detect illness early before it manifests with symptoms, so long as there is something we can do about it.
If you take a large group of people, there has to be some significant amount of hidden disease in that group for any medical screening test to make sense. If you take 5,000 people and only one person actually has a hidden disease, then it doesn’t make sense to test all 5,000.
This notion flies in the face of our frequent sense that no effort is too great when looking for hidden disease. Many people feel that more is always better. This is not true. Whether it is a skin test for tuberculosis, an electrocardiogram for heart disease, or bone density testing for osteoporosis, more testing of groups where there is very little disease can cause more problems than it prevents.
Every medical test is imperfect, and most have concrete risks. Every medical test has false positives (the test says you have a disease when you really don’t) and false negatives (the test says you are okay when you actually have the disease).
If a group of 5,000 people has very few individuals with the disease, say only one or two, then you are bound to get more false positives when you test them, because the tests are not perfect.
With breast cancer, there is a steady increase in cases as women get older. There are very few breast cancers in women in their 30s and 40s, so the false positive rate is higher. These women will be getting unnecessary M.R.I.s, biopsies, and other procedures. But don’t worry — you can still get a mammogram in your 40s if you want one. These are only guidelines.
“Screening mammography in women under 50 is problematic, and has always been problematic,” said Dr. Darrell N. Smith, a breast cancer detection expert and assistant professor of radiology at Harvard Medical School. “This is not a new debate. Ten years ago, we were debating whether we should screen at all under age 50,” he told me.
We need screening tools that have good sensitivity for detecting disease (low false negatives), and good specificity (the test detects an important disease specifically and not some other more trivial condition). Most important, you have to detect a disease that would matter to your medical outcome.
Detection of abnormal cell changes that might not ever advance to invasive cancer confuses things. Patients get upset by any “abnormality,” but it has to be put into context by your doctors. There are many changes in the breast that are not and will not become lethal cancer.
Likewise, if you are detecting a disease for which nothing can be done, then screening is useless. The screening test needs to be able to make a difference in outcomes. The disease should be treatable at that stage.
All guidelines must be based on the science, what we call evidence-based medicine. Otherwise we may cause more harm than good. Screening sounds like a great idea, but you must have the science to back it up.
People get emotionally invested in screening tests, but is it a false sense of security? For example, in the best radiology centers, about 90 percent of breast cancers can be detected with mammography. But of the women who are sent for biopsy, about 75 percent will not have breast cancer. So you see, we need to be more precise and thoughtful about our recommendations.
There are only so many breast radiologists in the United States, and if we screen women who have a low chance of having a treatable cancer we will divert resources from reading highly needed studies. But even if we had more resources, it would still not make sense to screen low-rate populations. This is not really rationing — it is rational.
Questions can be directed to Dr. James Dillard at jdillard@ehstar.com.