A patient limps into my office with knee pain. I thoroughly review her symptoms, treatments and past health problems. Then I examine her. Based on this information, I suspect a tear in the cartilage of her knee (called the meniscus). I recommend an MRI.
A few days later, I get the results: Sure enough, there's a small tear in the meniscus, in the same area as her pain. And yet, I am not sure the MRI results explain her symptoms. And by ordering the MRI, I may have wasted hundreds of dollars and put her on a path to surgery that she (and I) will regret. In fact, her abnormal MRI may have nothing to do with her pain.
Perhaps this doesn't make much sense. But it's true.
Back to top
To Test or Not to Test?
Perhaps it seems obvious that testing is the best next step for someone with significant symptoms and no clear diagnosis. But, there are reasons to think twice. I'm often faced with a dilemma in my practice for patients with symptoms: Do I order a test, treat without a test or do both? Each of these options has merit.
Here are some of the major pros and cons when it comes to testing. (These are the ones that I think are most important; perhaps you or your doctor can come up with others.)
- You may get a quick and definitive answer.
- The answer you get may point to the right therapy.
- The results of testing may predict the course of your condition.
- High tech images that may reveal the cause of your symptoms have a high "coolness" factor. They're amazing, irresistible and seductive. In fact, to some people testing is a foregone conclusion. Why wouldn't you just order the test?
- Testing may detect an unexpected and important abnormality unrelated to the reason the test was ordered.
- The results might point to additional testing (which, in turn, could offer the other benefits listed above).
- Healthy people often have false-positive results. That's when a test shows a problem when there isn't one. This can make it difficult to interpret results when a person has a disease.
- People with disease may have normal (or false-negative) results.
- Testing can be uncomfortable. Ask anyone who has had a mammogram, Pap test or prostate examination lately.
- Testing can be frightening. People who are claustrophobic may become quite distressed in an MRI scanner. Sedation may be necessary.
- Some tests come with potentially toxic exposures, such as dye or radiation.
- Testing can be risky. Even common or routine tests, such as colonoscopy or cardiac catheterization, are occasionally associated with serious complications.
- Waiting for test results can be associated with significant anxiety. It may be hard to appreciate this until you've been tested for HIV or scanned for cancer and then realize it's been four days and a long weekend is coming up and you still haven't received the results. Is the delay because the news isn't good?
- Test results may not give you any new information. For example, an X-ray may show swelling of the knee. But your doctor probably knew that already after examining you.
- Testing is not free. Whether you pay for all, part or none of the cost of a test, payment comes from somewhere. Ultimately we all bear the cost. Tests involving technology tend to be particularly expensive.
- Many conditions do not cause abnormal test results. Examples include fibromyalgia, migraine headaches and irritable bowel syndrome.
- Test results can be confusing or inconclusive. This can lead to more testing (beginning a cycle of additional testing with all of the problems above).
Back to top
Are Test Results Accurate?
Some tests are more accurate than others. But most tests are far from perfect. I think the notion that healthy people can have abnormal test results is underappreciated. Many of these abnormalities are meaningless. They are more of an indication of how sensitive our tests are than an indication of some "silent disease" lurking unsuspected in healthy people.
In people without symptoms, minor abnormalities occasionally predict future trouble, but at least as often, they cause unnecessary anxiety.
Common examples of test inaccuracies include:
- MRI of the spine to look for disc disease – According to studies, up to 85% of people without back symptoms have disc abnormalities such as degeneration or bulging.
- MRI of the knee to find a cause of pain – A sore knee often leads to MRI studies. However, a study published in 2008 found that 61% of people found to have at least one tear in their knee cartilage had no pain, aching or stiffness during the previous month.
- MRI of the brain – This can show findings suggestive of multiple sclerosis (MS) even in the absence of disease. In one study, 70% of people who had no symptoms but had brain MRI findings suggestive of multiple sclerosis (MS) did not develop the disease over more than five years.
These studies suggest that treatment based solely on MRI findings may not be helpful.
Of course, some tests are highly accurate. For instance, we're lucky that two excellent tests for HIV were discovered soon after the AIDS epidemic began. They're rarely wrong when it comes to screening and confirming a person's HIV status (though, as good as the HIV tests are, standard antibody testing for HIV just after infection can be falsely negative).
Back to top
What's a Person to Do?
Ask your doctor about the pros and cons of testing for your symptoms or condition. If you have questions - why a particular test has been recommended or why you aren't having any tests - ask for more information!
If your doctor does recommend a test, be sure you understand why it's being ordered and how the results might be helpful. Consider asking the following questions:
- Is the test useful for your condition? For example, does it provide accurate information about the presence or absence of a particular condition?
- Will the test affect treatment? For instance, an MRI of the spine looking for a herniated ("slipped") disc may not be helpful for a person who has mild symptoms or who has already decided that surgery is not something they would accept even if recommended. In such cases, it's unlikely that treatment would change based on the MRI results.
- When and how will you get the results? Knowing when to expect test results can reduce the anxiety that often accompanies the wait.
- What's the risk of not having a test? For many conditions, it's reasonable to begin a course of treatment or simply allow the passage of time before performing scans or other tests. If you improve with this approach, testing may be unnecessary.
Back to top
The Bottom Line
Patients sometimes ask me to "just test for everything." I tell them to be careful what they wish for! There are literally hundreds of tests that can be done and all of them have significant limitations. Many come with risks. All of them come with a cost.
Modern medicine has a lot to offer. The right test at the right time can provide a quick answer that directs treatment, speeds recovery and even saves lives. But testing is often associated with more harm than good. The trick is figuring out when tests will be helpful and when to do without them. Unfortunately, there's no test available that tells us when testing is most useful.
Back to top
Robert H. Shmerling, M.D. is associate physician at Beth Israel Deaconess Medical Center and associate professor at Harvard Medical School. He has been a practicing rheumatologist for over 20 years at Beth Israel Deaconess Medical Center. He is an active teacher in the Internal Medicine Residency Program, serving as the Robinson Firm Chief. He is also a teacher in the Rheumatology Fellowship Program.