How prevalent are false diagnoses of disease?

Recently, we expressed concern about the effects on the accuracy of the diagnostic process of the increasing numbers of well and worried well entering the medical care system.

One of the consequences of this influx of well people (and the concomitant reduction in disease prevalence) is the generation of more false positive test results and false diagnoses of nonexistent diseases.

The medical literature is filled with studies on the accuracy of specific disease diagnoses but the focus has been exclusively on missed diagnoses. These studies have often used autopsy data to discover how many patients died with specific diseases overlooked in life.

While missed diagnoses certainly deserve our attention, the opposite error has been almost completely ignored: How many patients with specific diagnoses of disease do not have the named disease? How prevalent are false diagnoses of disease? And which ones?

We are puzzled that these questions are not only unanswered but seem ignored in the literature.

We have found only one paper that looks at the amount of false diagnoses of a specific disease in a defined population.

In an article entitled “The Morbidity of Cardiac Non-Disease in School Children” on pages 1008-1013 of the NEJM‘s volume 276 in 1967, Bergman and Stamm reported on having studied over 100,000 school children in Seattle.

When asked, slightly over 100 reported having a diagnosis of heart disease. But when carefully evaluated, only 20% had heart disease and 80% did not. The measured physical and psychological disabilities were 75% in both groups.

Four times as much disability came from the false label “heart disease” as came from actual heart disease.

Obviously, these false diagnoses were harmful.

Why is the error of over-diagnosis so ignored? Consider a person who has been diagnosed with a nonexistent disease:

  1. The disease cannot progress since it does not exist.
  2. The patient is often satisfied to have a name, any name, for his or her problem, even if the disease does not exist.
  3. The doctor is satisfied to have named the condition, believing the diagnosis to be correct, and
  4. The false positive test result leading to the diagnosis may become negative when repeated in the future, reassuring the patient and the doctor that the “disease” is in remission or is a mild form.

Unless the patient sees another doctor who doubts the false diagnosis and repeats the testing, the false diagnosis will persist.

Michael Balint studied family physicians in the 1950s and wrote a book called “The Doctor, His Patient and The Illness.” He concluded that once a doctor and a patient agreed on a diagnosis, the “non-disease” becomes incurable.

Our 50 years in medicine tells us that this error is common and that it is very difficult to remove a false diagnosis of any disease.

George Lundberg is a MedPage Today Editor-at-Large and former editor of the Journal of the American Medical Association and Clifton Meador is an endocrinologist on the faculties of Vanderbilt and Meharry Medical Schools.

Originally published in MedPage Today. Visit for more health policy news.

View 3 Comments >

Most Popular

✓ Join 150,000+ subscribers
✓ Get KevinMD's most popular stories