Diagnosis and undiagnosis

Neurologists often pride themselves on their diagnostic skills (figure 1). Indeed, we often assume that having made a firm diagnosis, the flow of initial treatment and longer-term management will follow almost automatically. But with long-term review of patients, neurologists are perhaps better than most at challenging the so-called ‘established diagnoses’; not infrequently they question a previously assigned diagnostic label, or even find it necessary to remove a diagnosis from a patient. Removing a diagnosis lacks a descriptive verb; euphemisms might include ‘maturation of the diagnosis’, ‘diagnostic refinement’ and so on. We suggest that the more robust term ‘undiagnosis’. Stone et al1 have discussed the different degrees of diagnostic shift. Here we address specifically the process of dealing with previous diagnostic errors, imparting a new diagnosis and removing a previous one.

The process of ‘undiagnosing’ has not been studied in any detail; some might argue that it does not merit detailed study. Yet, the impact of the removal of a long-held diagnosis—not just on the patient, but also on their family, carers and other health professionals—may be substantial. Some might assume that removing an erroneous diagnosis could only be a positive event for a patient, but our experience suggests otherwise. We have witnessed the creation of considerable distress, anger and frustration in individuals who may have shaped their lifestyle, work pattern and home life according to a neurological label attached in years gone by.

Advances in diagnostic techniques seem likely to increase the need for neurological undiagnosis. Experienced practitioners will recognise the scenario of the patient with presumed Parkinson's disease whose dopamine transporter scan remains steadfastly normal, or the patient whose diagnosis of multiple sclerosis stretches back over several decades but whose first MRI scan is pristine in its normality. Future undiagnoses …