Intended for healthcare professionals

Reading, Writing, And Revalidation

What should we say to patients with symptoms unexplained by disease? The “number needed to offend”

BMJ 2002; 325 doi: (Published 21 December 2002) Cite this as: BMJ 2002;325:1449
  1. Jon Stone, research fellow in neurology (jstone{at},
  2. Wojtek Wojcik, medical studenta,
  3. Daniel Durrance, medical studenta,
  4. Alan Carson, consultant neuropsychiatristb,
  5. Steff Lewis, medical statisticiana,
  6. Lesley MacKenzie, sister in neurology outpatientsa,
  7. Charles P Warlow, professor of medical neurologya,
  8. Michael Sharpe, reader in psychological medicineb
  1. a University Department of Clinical Neurosciences, Western General Hospital, Edinburgh EH4 2XU
  2. b University Department of Psychiatry, Royal Edinburgh Hospital, Edinburgh EH10 5HF
  1. Correspondence to: J Stone

    Most doctors make a diagnosis and offer treatment to patients whose symptoms turn out to be unexplained by disease.1 In such cases a diagnostic label is important in signifying to the patient and family that the doctor is taking the problem seriously and accepts the complaints as real. Some diagnostic labels, particularly those that sound “psychological,” can be perceived by patients as offensive by implying that the patients are “putting on” or “imagining” their symptoms or that they are “mad.”2

    Various potentially suitable diagnoses are available to doctors. “Hysteria” was the traditional term and is still sometimes used. “Functional nervous disorder” was used in the late 19th century to denote symptoms arising from disordered nervous functioning,3 but in the 20th century this was superseded by terms that implied psychogenesis, such as psychosomatic.4 In the past 20 years more neutral descriptive terms such as “medically unexplained symptoms” have gained in popularity.1

    Despite their importance in the doctor-patient relationship, the implications to patients of these labels have received remarkably little attention. We explored the differing connotations and potential offensiveness of 10 different medical labels for the symptom of weakness.

    Participants, methods, and results

    The study received local research ethics approval. Two medical students (WW and DD) interviewed 86 consecutive new patients attending a general neurology outpatient clinic in Edinburgh, before patients saw the doctor. Twenty four other patients declined to take part (most because they were in a hurry), and three further interviews were incomplete. We asked patients, “If you had leg weakness, your tests were normal, and a doctor said you had [diagnosis] X, would he or she be suggesting [implication] Y?” The table shows the 10 diagnostic labels for weakness (X) and five potential connotations (Y). We coded patients' responses “yes,” “no,” or “don't know” for each diagnosis and each connotation.

    ”If you had leg weakness, your tests were normal, and a doctor said you had ‘X’ would he be suggesting that you were Y (or had Y).” Percentage responses among 86 new neurology outpatients, offence score, and “number needed to offend”—that is, number of patients who would have to be given this diagnostic label before one patient is “offended”

    View this table:

    The diagnoses of multiple sclerosis and stroke always had fewest negative connotations and “symptoms all in the mind” the most. The diagnoses ranked in between were of greater interest. We calculated an “offence score” for each diagnosis as the proportion of patients who endorsed one or more of the following connotations, which we deemed offensive: “putting it on,” being “mad,” or “imagining symptoms.” We then used this value to calculate a “number needed to offend”—that is, the number of patients who can be given this diagnosis before one patient is offended (see figure on This value assumes an ideal world in which no one is ever offended, and we used standard calculations for number needed to treat.5 A comparison of “medically unexplained weakness” and “functional weakness,” two of the most popular labels in use, revealed that “functional” was much less offensive (P<0.05 for all categories of negative connotation, McNemar's test).


    Many diagnostic labels that are used for symptoms unexplained by disease have the potential to offend patients. Although “medically unexplained” is scientifically neutral, it had surprisingly negative connotations for patients. Conversely, although doctors may think the term “functional” is pejorative,6 patients did not perceive it as such. As expected, “hysterical” had such bad connotations that its continued use is hard to justify, although it is the only term in this list that specifically excludes malingering.

    Diagnostic labels have to be not only helpful to doctors but also acceptable to patients. Many of the available labels did not pass this basic test, but “functional” (in its original sense of altered functioning of the nervous system3) did. This label has the advantage of avoiding the “non-diagnosis” of “medically unexplained” and side steps the unhelpful psychological versus physical dichotomy implied by many other labels. It also provides a rationale for pharmacological, behavioural, and psychological treatments aimed at restoring normal functioning of the nervous system.4 We call for the rehabilitation of “functional” as a useful and acceptable diagnosis for physical symptoms unexplained by disease.


    Contributors: JS developed the study and discussed core ideas with AC, MS, and CPW. WW, DD, SL and LM collected data and participated in data analysis. All authors contributed to writing the paper. MS is the guarantor.


    • Funding None.

    • Competing interests None declared.

    • Embedded Image A figure appears on


    View Abstract