Why Doctors Order Too Many Tests

While cleaning out some old files, I was delighted to find an article I had clipped and saved 35 years ago: a “Sounding Boards” article from the January 25, 1979 issue of The New England Journal of Medicine. It was written by Joseph E. Hardison, MD, from the Emory University School of Medicine; it addresses the reasons doctors order unnecessary tests, and its title is “To Be Complete.” Today we have many more tests that can be ordered inappropriately and the article is even more pertinent and deserves to be re-cycled. He says,

When challenged and asked to defend their reasons for ordering or performing unnecessary tests and procedures, the reasons given usually fall under one of the following excuses…

He lists ten excuses:

  1. To be complete. (This implies a need to go beyond a rational diagnostic process. “Complete” is a nebulous concept; you could always do one more test.)
  2. The “they say” excuse. (Who are “they” and do they really say that? If they do, do they say it for science-based reasons or is it just an opinion?)
  3. The “we’ll get in trouble if we don’t” excuse. (This expresses student doctors’ fears of the attending physician, who may not actually chastise them for not ordering an alpha-fetoprotein test on every patient.)
  4. The “if we don’t order everything at once, it won’t get done” excuse, commonly given in large city hospitals. (Should rational health care suffer because of institutional inefficiency?)
  5. The “as long as he is in the hospital, we might as well” excuse. (Why should a patient’s care depend on where he’s located or how readily available a test is?)
  6. The “academic” excuse. The false idea that the evaluation of a patient should be somehow different or more complete in an academic institution. Here Dr. Hardison quotes a dictionary definition of an academic as “very learned but inexperienced in or unable to cope with the world of practical reality.”
  7. The “malpractice” excuse. I have personally heard this excuse all too many times. One of my friends was challenged on the witness stand about why he ordered an x-ray in a trauma case when he had testified that it was not really indicated by the patient’s findings. He answered, “I ordered it so I could tell you here in court today that I had done so.” It’s a sad state of affairs when doctors make decisions based on fear of malpractice suits rather than on the patient’s welfare.
  8. The “protocol” excuse: the patient is a candidate for a study that requires these tests for its protocol. (Best find out first if the patient is really going to be enrolled in the study and consult the researchers or wait and let them order the tests they need.)
  9. The “if it were my mother or father, I’d want it done” excuse. He comments, “The parent who has his child for a doctor had best beware.” (In other words, when doctors are emotionally invested in patients, reason flies out the window.)
  10. The “how do we know he doesn’t have it?” excuse. (Carrying that excuse to its logical consequences would mean ordering every possible test on every patient, and getting a lot of false positive and irrelevant results. A better question is “do we have any reason to suspect he might have it?”)
  11. I can think of two more excuses:

  12. The “knowledge is good” excuse. If anything is abnormal, we want to know. The more we know about our bodies, the better. (That’s not always true; sometimes we are better off not knowing irrelevant or insignificant information.)
  13. The “fishing expedition.” I don’t have any idea what’s wrong with this patient, but maybe if I order a lot of tests, something will turn up. (Not only is this an admission of incompetence, it’s likely to turn up misleading information, to lead to the wrong diagnosis, and to do more harm than good.)

Too many tests can be hazardous to your health for several reasons:

  • Because of the way normal values are determined (testing a lot of presumably normal people and cutting off the ends of the resulting Bell curve), there is a good chance that one out of every twenty tests will give an “abnormal” result that is not really abnormal. Being 6 feet 7 inches tall falls at one extreme of the Bell curve but it is “normal” for that person and doesn’t mean he has a disease.
  • False positive results become increasingly likely when a disease is rare and/or when the patient’s history and physical don’t already point to that diagnosis.
  • Following up on false positive results can be a wild goose chase, with more unnecessary tests and procedures. There’s no way to tell for sure whether a suspicious shadow on an x-ray represents a deadly disease or a harmless artifact without invasive procedures that carry risks.
  • Imaging procedures and other tests frequently identify “incidentalomas,” abnormal findings that are mere curiosities and that have no impact on the patient’s health other than to sometimes cause unnecessary worry.
  • Some tests involve potentially harmful radiation; and even the simple act of drawing blood can cause pain, bruising, and a tiny risk of infection. Even minor risks are not justified if the likelihood of benefit is too low.
  • Lab errors occur: machine malfunctions, misreading results, mixing up samples, recording and transcription errors, etc.
  • Overdiagnosis leads to unnecessary treatments.
  • Tests cost money, sometimes a lot of money, which is not healthy for your own wallet or for society’s health care budgets.

The need for a test can be informed by scientific studies. Does routinely ordering x-rays on all patients with ankle injuries improve outcomes? No, it doesn’t. Simple sprains are much more common than fractures, and x-rays expose patients to radiation. Science-based guidelines like the Ottawa ankle rules have been developed to help clinicians decide when to order tests.

Another consideration is “what difference will the test make?” What are we going to do differently if the result is x rather than y? If we can’t answer that question, we probably shouldn’t be doing the test. That’s particularly pertinent in genomic testing, where patients may be told they are at high risk of developing a disease that they can’t do anything to prevent.

And then there’s CAM. Many tests offered by CAM practitioners have not been validated, some are known to be bogus, and some are used to diagnose bogus diseases.

And then there are the patients who demand tests because of something they read on the Internet.


Every year there are more tests available for doctors to order. Doctors should not order any of them without good reason. Doctors should be guided by good judgment grounded in science. Patients should not be hesitant to question their doctors if they don’t understand why a test is being done or what difference the results will make.

Posted in: Diagnostic tests & procedures

Leave a Comment (128) ↓