Intended for healthcare professionals

Editorials

Telling patients there is nothing wrong

BMJ 1996; 313 doi: https://doi.org/10.1136/bmj.313.7053.311 (Published 10 August 1996) Cite this as: BMJ 1996;313:311
  1. Ray Fitzpatrick
  1. Professor Department of Public Health and Primary Care, Radcliffe Infirmary, University of Oxford, Oxford OX2 6HE

    Unless their true fears are addressed, diagnostic tests may leave them more anxious than before

    Reassuring patients who have unwarranted concerns that they are seriously ill is one of the commonest medical tasks. But the dearth of research on the subject suggests either that dealing with such patients is not a problem or that it is one best left to clinical judgment. In this issue of the BMJ (p 329), McDonald and colleagues report that many patients continue to be anxious about their heart despite being informed of a normal result from echocardiology.1 Three quarters of the patients in their sample were referred for exclusion of heart disease after routine examination for insurance or employment, suggesting that unjustified concern about ill health is often iatrogenic.

    Other studies have drawn attention to difficulties in reassuring patients. Mayou reports from a prospective study that one third of patients in a cardiac clinic who were told that there was nothing wrong with their hearts expressed fears about cardiac disease three months later.2 In a similarly designed study, 40% of a sample of patients referred to neurologists for benign headache expressed concern that their symptoms reflected serious disease one month after receiving what the neurologists termed “reassuring” results from the specialist.3 Patients receiving investigations such as skull radiography were no more likely to be reassured.

    Investigative tests may produce positive psychological benefits. Sox and colleagues conducted a randomised trial among patients with chest pain considered unlikely to be due to heart disease.4 Patients were all given reassurance on the basis of their history but only one group was randomised also to receive an electrocardiogram and blood test. Two thirds of patients were concerned about serious disease at baseline. Three weeks after the index visit only 20% expressed such concerns; a similar proportion was assessed at four months. Those receiving investigative tests were no more likely to be reassured. However, patients in this group were more satisfied with their care and made a speedier recovery in terms of disability arising from chest pain. The authors speculate on psychological mechanisms for this effect of diagnostic tests. One year after attending neurological clinics for headache, those receiving investigative tests at baseline experienced greater improvement in their symptoms independently of other aspects of management.5

    The usual culprit for failure to reassure is poor communication. A recent survey of 5150 randomly chosen patients who had recently been discharged from hospital found that a worrying 34% had not been told the result of tests.6 Kessel argues that to reassure a patient, appropriate and relevant information must be given.7 Merely asserting that there is nothing wrong can seem to deny the reality of the patient's concern. He also stresses rapport; reassurance is more likely when patients feel that the doctor has understood them as individuals. Explicitly addressing the presence of fears can also help. Some evidence supports Kessel's principles. A study of family doctors in Ontario looked at independent measures of doctors' patient centred communication styles, with particular emphasis on whether, in the opinion of an observer, the patient was able to express thoughts and feelings in relation to the presenting problem.8 In research assessments, over half of patients expressed substantial concern about their symptoms before the consultation, of whom 71% experienced subsequent reduction in concern. This reassurance was significantly more common in those patients who experienced patient centred consultations.

    Patients with substantial psychiatric symptoms are more difficult to reassure, as are those with a syndrome of hypochondriacal concern that may exist independently of depression.9 McDonald and colleagues go beyond such psychiatric thinking to refer to a variety of chance “wild card effects” that can leave patients resistant to reassurance—for example, seeing a graphic portrayal of death from a particular disease in the media.1 Fifty seven per cent of patients with worries about serious illness in the study of neurological management of headache had no psychiatric symptoms.3 5 Their concerns arose from a normal reaction to a combination of unusual circumstances—for example, patients experiencing alarming headaches at the same time as hearing of a neighbour's death from a brain tumour.

    Potentially adverse psychological effects of screening healthy people for cardiovascular risks have received some attention.10 We may need to evaluate diagnostic procedures in the same way. A recent study charts the rapid rise of diagnostic tests such as computed tomography and magnetic resonance imaging of the spine and cardiac catheterisation in the United States. It argues that rates of subsequent therapeutic interventions such as coronary artery bypass grafting and back surgery can be largely explained by this increased use of diagnostic technology.11 The authors warn of adverse consequences of the “increased sense of vulnerability associated with the transformation of a person into a patient” that may arise from this expansion.11 They largely have in mind those found to have an abnormality.

    As McDonald and colleagues argue, informing patients of a normal diagnostic test result may not always succeed in reassuring the patient. Ambiguous or false positive test results may also create or reinforce anxiety. Direct discussion of patients' concerns is more likely to reassure patients than unnecessary further tests or specialist referrals.

    Reference

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.