Visual hallucinations in Parkinson's disease: their nature, frequency, and origins.

The question mark in the heading of this series is deliberate. The most confidently applied diagnoses used for neurological disease are perhaps those in which a reasonably stereotypical clinical presentation has a known pathological substrate, in turn the consequence of a specific biochemical or enzymatic dysfunction. McArdle’s disease springs to mind as one such example, although even there some clinical variability is encountered in the expression of myophosphorylase deficiency. More uncertainty in diagnosis appears when the disease in question shows a much greater degree of variability in its clinical expression, and where the capacity to confirm the pathological substrate is limited. The diVerences between primary progressive multiple sclerosis and remittingrelapsing multiple sclerosis are profound, embracing clinical expression, imaging characteristics, and pathological features. Despite that, the entities remain under the same diagnostic umbrella. The greatest uncertainty arises when diseases have been described, often eponymously, on the basis of a few cases, and with restricted access to imaging or to pathological characterisation. Inevitably re-examination of these conditions—for example, Schilder’s disease—indicates that the case material is not homogenous and that the use of such titles, implying a uniform pathological mechanism, is best abandoned. The 13 articles in this series, look at various clinical entities, or diagnoses, the specificity of which remains open to question. In some instances, the whole concept of the entity is perhaps in doubt—for example, the whiplash syndrome. In others, a clinical syndrome—for example, the Tolosa-Hunt syndrome—may in reality represent the end state of heterogenous pathological states. The authors of this series have been asked to analyse the concept of the individual syndromes, determining whether they remain of value in neurological practice—indeed, whether the question mark can be sensibly removed, cautiously retained, or the diagnosis, as an entity, abandoned altogether.